- Imputation of missing values
- Additional classifiers such as
XGBoost
, Decision Trees - Better internal code structure
- Lot more tests
- More precise tests, as we vary number of classes wildly in test suites
- several bug fixes and enhancements (more cmd line options such as
--print_options
)
- new classifier : SVM
- new flag to choose a feature selection method
- user chosen options now saved to disk, to better handle complex interactions between options
- code clean up and faster tests
- Parallelizing the main the CV loop, leading to great reduction in total time for report generation!
- More options, including choice of different classifiers (Random Forest and Extra Trees classifiers)
- Support for dataset in Weka's ARFF format
- Better visualizations (handling small/nan values in feature importance, layouts and design)
- auto versioning!
- Ability to read meta data from pyradigms or ARFF files, without having to specify that separately.
- Dropping support for Python 2.7 :(